J.Org. Chem., Vol. 41, No. 3, 1976 557
Notes
-
Table I Conversion of Acyl Cyanides to 2-Substituted 3,3-Dichloroacrylonitriles ______
C1, C=CRCN, % yield CCl, a method
RCOCN, R
61 45c 40C 48 20 60
Ia, CH, Ib, CH, CH, I C , (CH,), CH Id, (C%)3C Ie, p c l C , H, If, p-CH, OC, H,
CCl, Br b 90 82 87 91
43 85
a The CCl,, triphenylphosphine procedure is reported in ref 4 and 5. b Yield based o n pure distilled or recrystallized
product.
C
Yield based o n VPC analysis of crude distillate.
ucts and lower yields result whenever more sensitive carbonyl components are added.4.5 We now wit3h t o report on a n improved method for t h e preparation of I1 a n d some reactions of I1 with acyl and aroyl cyanides. It has been found that t h e substitution of bromotrichloromethane for carbon tetrachloride allows the Wittig reagent I1 t o be formed much faster a n d t h e concurrent reaction with the carbonyl component is significantly improved with higher yields and virtually n o by-product formation. Gas chromatographic analysis of the crude reaction mixture does not show a n y of t h e bromochloroethylene derivative. Although this procedure should be applicable toward the conversion of other carbonyl containing compounds to the dichlorovinyl group (=CClz) we have studied t h e carbonyl of acyl cyanides so that yields may be directly compared with our previous work. 2Ph3P
+ C13CBr
-
RCOCN t Ph3P=CC12
+ PhaPBrCl Ph3PO + C12C=CRCN
Ph3P=CClz --*
I n a typical reaction, excess d r y bromotrichloromethane was added to 1.0 equiv of triphenylphosphine in sufficient d r y benzene t o produce a convenient volume. T h e mixture was stirred under a d r y nitrogen atmosphere at Oo for 15-30 min, then 0.5 equiv of t h e acyl cyanide, Ia-f, was added and t h e mixture stirred at 0' for 2-4 hr. An immedia t e shift of t h e nitrile band as well as appearance of a strong =CC12 band in t h e 9 2 0 - 9 4 0 - ~ m - ~region was observed7. After approximately 30 min, large quantities of triphenylphosphine oxide began t o precipitate from t h e reaction mixture. After 1-2 hr, t h e carbonyl stretching band was almost totally gone. Stirring the mixture for a longer period of time had little effect on the yield. I n cases using aliphatic acyl cyanides, the reaction mixture was rapidly warmed and benzene, excess bromotrichloromethane, and liquid products were removed under vacuum. Unlike our previous r e p ~ r tt,h~e 2-alkyl-3,3-dichloroacrylonitriles were virtually free of impurities and easily purified by simple distillation t o give yields of 82-91% (Table I). With aroyl cyanides, the benzene and excess bromotrichloromethane was stripped from t h e reaction mixture a n d crude products extracted with boiling ligroin a n d t h e crude products recrystallized from anhydrous methanol t o give yields of 43-
85%. Experimental Section All reactions were carried out under a dry nitrogen atmosphere. Boiling points and melting points are uncorrected. Infrared spectra were performed on a Beckman IR-8 and calibrated at 1601.0 cm-' with polystyrene film. Commercial triphenylphosphine was used without purification. The bromotrichloromethane and ben-
zene were purified by atmospheric distillation, using only the center cuts. Acyl Cyanides. The acyl cyanides used in this study were prepared according to procedures reported p r e v i o u ~ l y . ~ ~ ~ Reaction of I1 with Aliphatic Acyl Cyanides. General Procedure. A mixture of triphenylphosphine (0.15 mol) and 80 ml of dry benzene was cooled in an ice bath and then 40 g of bromotrichloromethane was added. The mixture was stirred at Oo for 30 min and then 0.075 mol of the aliphatic acyl cyanide (Ia-d) was added rapidly. After maintaining the reaction at Oo for approximately 3 hr it was heated under vacuum (25-30 mm) to yield benzene, the excess bromotrichloromethane, and crude product. A second distillation of the crude product gave pure 2-alkyl-3,3-dichloroacrylonitriles with physical properties and ir spectra identical wi$h those reported previ~usly.~ Reaction of I1 with Aryl Acyl Cyanides. General Procedure. In a manner similar to that described above, the aryl acyl cyanide (Ie,f) was added rapidly to a mixture of benzene, triphenyiphosphine, and bromotrichloromethane a t 0'. After 3 hr, the mixture was vacuum stripped to remove the benzene and excess bromotrichloromethane. The solid residue was extracted three times with 100-ml portions of boiling ligroin. The combined extracts were cooled to remove most of the triphenylphosphine oxide. Concentration and cooling the extracts gave a crude product which was recrystallized twice from anhydrous methanol. The pure 2-aryl3,3-dichloroacrylonitriles displayed physical properties and ir spectra identical with those reported previously?
Acknowledgment. T h e support of this research by t h e Robert A. Welch Foundation is gratefully acknowledged. Registry No.-Ia, 631-57-2; Ib, 4390-78-7; IC, 42867-39-0; Id, 42867-40-3; Ie, 13014-48-7; If, 14271-83-1; 11, 6779-08-4;triphenylphosphine, 603-35-0; bromotrichloromethane, 75-62-7; 2-methyl3,3-dichloroacrylonitrile,31413-58-8; 2-ethyl-3,3-dichloroacrylonitrile, 42791-06-0; 2-isopropyl-3,3-dichloroacrylonitrile, 42867-43-6; 2-tert-butyl-3,3-dichloroacrylonitrile, 42867-44-7; 2- (p-chlorophenyl)-3,3-dichloroacrylonitrile, 37447-52-2; 2-(p-methoxyphenyl)3,3-dichloroacrylonitrile,37447-53-3.
References and Notes (1) A. J. Speziale and K. W. Ratts, J. Am. Chem. SOC.,84, 854 (1962). (2) R. Rablnowitz and R. Marcus, J. Am. Chem. Soc., 84, 1312 (1962). (3) C. Raulet and E. Levas, C. R. Acad. Sci., Ser. C, 270, 1467 (1970). (4) E.A. Clement and R . L. Soulen, J. Org. Chem., 39, 97 (1974). (5) R. L. Soulen, S. D. Carlson, and F. Lang, J. Org. Chem., 38, 479 (1973). (6) D. Seyferth, H. D. Simmons, Jr., and G. Singh, J. Organomet. Chem., 3, 337 (1965). (7) R. L. Soulen, D. B. Clifford,F. F. Crlm, and J. A. Johnston, J. Org. Chem., 36, 3386 (1971).
Synthesis of Some New Hindered Biaryls Harry W. Gibson* and F. C. Bailey Webster Research Center, Xerox Corporation, Webster, New York 14580 Received June 11.1975 Our studies of the effects of chemical structure on molecular orbital energy levels a n d electron d i s t r i b ~ t i o n s l -and ~ associated electrical properties6 have concerned themselves primarily with planar aromatic systems. Some such systems are subject t o steric effects or nonbonded interactions which can lead to deviations from coplanarity and hence deviations from anticipated behavior. T h e biphenyl system is one whose susceptibility t o these effects is well known.7 T h e target structures 1-5 were designed with a view toward providing (1) a high rotational barrier about the biphenyl bond and (2) systems with a n electron-rich ring and an electron-poor ring as well as ones with both rings electron poor or rich.
558
J . Org. Chem., Vol. 41, No. 3, 1976
Notes oil which could not be purified so as to afford satisfactory carbon analysis, it was derivatized. Condensation of 2, R = n-CsH17, with p-cyanobenzaldehyde afforded ani1 13 in 99% yield.
HzN+oR C1 CH,
C1 CH3
1
2
Br C1
I 13
4
3
CH, H,C CH, 5 Discussion
A. S e r i e s 1. The prime starting material in the synthesis of system 1 was 2,3,5-trimethylphenol (isopseudocumenol, 6), which was converted to the known8 7 in 99% yield. Formation of the previously unreported benzyl ether 8 (85% yield) was accomplished by treatment of 7 with benzyl chloride in alkaline solution. Attempts to benaylate 7 without isolation from the reaction mixture were less efficient. Similarly the new n-octyl ether 9 was prepared in 97%yield using n-octyl iodide and 7. A Sandmeyer reaction of diazotized 10 afforded a 71% yield of the desired 2-chloro-4-nitroiodobenzene ( I lL9
I
6
I 7,R=H 8, R = CHzC,H,
NO* 10, X = NH,
11, X = I
9, R = n-C8Hl, Cl
*
c1 12
An Ullmann reaction of equimolar amounts of 8 and 11 led to the isolation of the desired 4-benzyloxy-2’-chloro-4’nitro-2,3,6-trimethylbiphenyl (1, R = C H z C & , ) in 36% yield (17% conversion) by means of column chromatography. Accompanying 1, R = CHzC&, was 2,2’-dichloro4,4’-dinitrobiphenyl (l2).9 The desired 1, R = H, was obtained in 88% yield by refluxing 1, R = CHzCsHb, with a 30% solution of HBr in acetic acid. Formation of octyl ether 1, R = n-CEH17, was accomplished (89%)by treatment of 1, R = H, with octyl iodide in alkaline medium. 1, R = nCBH17, can be obtained in a more direct manner via an u l l mann reaction of 9 and 11. This transformation proceeded in 53%yield (17% conversion). B. S e r i e s 2. The nitro group of 1, R = n-CsH17, was reduced chemically with stannous chloride, affording a 97% yield of amine 2, R = n-C8H17. Since 2, R = n-CsH17, is an
14
C. System 3. The approach to series 3 involves utilization of one of the starting materials used for series 1, namely the benzyl ether 8. The other starting material is of structure 14, a 3 nitro-4-iodobenzoate ester. Nitration of ethyl p-iodobenzoate gives 14, R = CzH5.l’ 9 and 14, R = CzH5, were subjected to the conditions of the Ullmann reaction; by column chromatography 3, R = CzH5, R’ = nC8H17, was isolated in 8% yield. The ester was hydrolyzed to the desired acid 3, R = H, R’ = n-C8H17, in 90% yield. D. S e r i e s 4. Treatment of 12 with bromine in the presence of silver sulfate and sulfuric acidll led to the isolation of only dibromo derivative 4 in quantitative yield. Attempted brominations using organothallium intermediates were fruitless; apparently the aromatic rings of 12 are too deactivated for reaction with these species. E.S e r i e s 5. Subjection of 8 to conditions of the Ullmann reaction for normal reaction times led primarily to recovery of starting material. Extended reaction times (60 hr) caused charring-no identifiable products were isolated. The procedure was that of previous workers who had succeeded in coupling 2,6-dimethyliodobenzene to form 2,2’,6,6’-tetramethy1biphenyl.l2 The present result is somewhat surprising, but can be explained by the presence of t h e deactivating benzyloxy group and increased steric hindrance a t the 1-(iodo) position owing to the buttressing effect of the 3-methyl substituent. As an alternative to the Ullmann synthesis of 5, R = C H z C & , , a method previously employed for the synthesis of bimesityl13 was utilized. The Grignard reagent derived from 8 was treated with cupric chloride in anhydrous ether. The hydrolysis product, 3-benzyloxy-1,2,5-trimethylbenzene, which was also synthesized from isopseudocumenol (6),was isolated in 80% yield. This result in light of the synthesis of bimesityl is surprising; it can be rationalized only in terms of the buttressing effect of the 3 and 4 substituents of the Grignard reagent prepared from 4-benzyloxy-2,3,6-trimethyliodobenzene.This apparently makes the 1position very sterically hindered, even to a greater extent than in that derived from iodomesitylene (2,4,6-trimethyliodobenzene), so that the Grignard reagent does not couple. Thus, attempted syhtheses of 5, R = CHzCsH5, were unsuccessful. E x p e r i m e n t a l Section General. Melting points were taken in capillaries on a ThomasHoover apparatus and are corrected. Infrared (ir) spectra were recorded on a Perkin-Elmer Model 137; nuclear magnetic resonance (NMR) spectra were recorded on a Joelco C-60H unit and all shifts are relative to internal tetramethylsilane. Elemental analyses were performed by Spang Microanalytical Laboratories, Ann Arbor, Mich. 4-Iodo-2,3,5-trimethylphenol(7). This compound, mp 110.0110.5O, was prepared from 2,3,5-trimethylphenol (6) by the method of Cressman and ThirtleE using hydrogen peroxide-iodine in 98% yield: reported8 mp 112-113’; yield 88% NMR (CD3COOD) S, 6 2.20 (3 H, 2-CH3); s, 2.34 (3 H, 3-CH3); s, 2.42 (3 H, 5-CH3); s, 6.70 (1H, 6-H); s, 11.1 (1H, OH); ir (KBr) 3300 cm-l (OH). 4-Benzyloxy-2,3,6-trimethyliodobenzene(8). A solution of
Notes 63.4 g (0.243 mol) of phenol 7, 19.5 g (0.348 mol) of 87% potassium hydroxide, 670 ml of ethanol, and 31.4 g (0.248 mol) of benzyl chloride was refluxed overnight, concentrated by distillation of 400 ml of ethanol, and then poured into 1200 g of ice. After filtration the solid was washed with 5 M sodium hydroxide and water, then dried and recrystallized from ethanol to give 62.8 g (73%)of material, mp 69-74’. Two more recrystallizations gave colorless crystals, mp 74.0-716.0’. Anal. Calcd for C16H1710: C, 54.56; H, 4.87. Found: C, 54.67; H, 4.89. NMR (CDC13) s, 6 2.26 (3 H, 3-CH3); s, 2.44 (3 H, 2-CH3); 9, 2.46 (3 H, 6-CH3); S, 4.95 (2 H, OCH2); 9, 6.69 (1 H, 5-H); s, 7.30 (5 H, aromatic); ir (KBr) 1110, 1230, 1260 cm-I (ArOCH2-), no OH. 4-0ctyloxy-2,3,6-trimethyliodobenzene(9). A solution of 21.0 g (0.0800 mol) of phenol 7, 19.2 g (0.0800 mol) of n-octyl iodide, 5.87 g (0.0900 mol) of 87% potassium hydroxide, 225 ml of ethanol, and 100 ml of dioxane was refluxed for 18 hr. Then most of the ethanol was rernoved and the resultant mixture was poured into ice-water. The hexane extract was dried and reduced in volume, whereupon 10.9 g (36%) of solid, mp 29-30’, precipitated. The filtrate was elutedl through a column of neutral alumina and yielded 11.2 g (37%) more of the same material. Two recrystallizations from methanol--hexane yielded colorless crystals, mp 30.0-31.0’. Anal. Calcd for C17H27IO: C, 54.55; H, 7.27; I, 33.91. Found C, 54.72; H, 7.18; I, 34.03. NMR (CDC13) m, 6 0.7-2.18 (15 H, OCC7H15); 9, 2.21 (3 H, 3-CH3); 9, 2.45 (6 H, 2-CH3, 6-CH3); t (J = 6 Hz), 3.88 (2 H, OCH2C); s, 6.63 (1 H, 5-H); ir (KBr) 1050, 1230, 1270 cm-I (ArOCHZ),no OH. Activated Copper Powder. Copper powder was prepared by reduction of copper sulfate with zinc14 and stored under water. To “activate” it, 114 g of wet copper powder was stirred in a solution of 30 g of iodine in 200 ml of acetone. After washing with acetone it was dried in a vacuum oven to give 67.7 g of catalyst. This “activation” process was carried out just prior to each Ullmann reaction. 4-Benzyloxy-2~-chloro-4r-nitro-2,3,6-trimethylbiphenyl (1, R = CHzCsH5). An intimately ground mixture of 40.5 g (0.115 mol) of 4-benzyloxy-2,3,6-trimethyliodiobenzene(8) and 32.5 (11)9 was charged to a (0.115 mol) of 2-chloro-4-nitroiodobenzene 400-ml stainless steel beaker containing 1 g of activated copper powder and 2 ml of pyridine and heated with stirring in an oil bath to 220’ and a total of 66.7 g of activated copper was added at a rate of approximately 5 g/min. After heating for 1.5 hr, the mixture was cooled and exhaustively extracted with benzene. Concentration of the extract gave 37.8 g of a dark brown oil. This was chromatographed on neutral alumina with 1:l benzene-hexane to yield 12.4 g of recovered 8 and 7.3 g of desired biphenyl 1, R = CH&&&, mp 155-160’. Elution with 3:l benzene-methanol gave 7.54 g of 12 (see below). Based on unrecovered starting material 11, the yield and conversion for 1, R = CHzC&I5, are 36 and 54% and for 12,17 and 54%. After three recrystallizations from ethanol-acetone the pale yellow crystals of 1, R = CHZC&F,, had mp 165.0-166.0”. Anal. Calcd for CzzHzoClN03: C, 69.20; H, 5.28; N, 3.67; C1, 9.29. Found: C, 69.74; H, 5.36; N, 3.72; C1, 9.36. NMR (CDCl3) s, 6 1.97 (3 H, 3-CH3); S, 2.01 (3 H, 2-CH3); 2.31 (3 H, 6-CH3); S, 5.13 (2 H, OCH2); s, 6.86 (1 H, 5-H); m, 7.5 (6 H, 6’-H, C H Z C ~ H ~m,) ;8.4 (2 H, 3’,5’-H); ir (KBr) 1520, 875 (NO2); 1130, 1230, 1260 cm-‘ (ArOCH2-). 2,2’-Dichloru-4,4‘-dinitrobiphenyl(12). The crude product (above) after recrystallizations from benzene had mp 107.0-108.5” (reportedg mp 107’1, yield 44% from 11. Using the same procedure as for the preparation of 1, R = CH&&I5,25.0 g (0.0877 mol) of 11 and 18 g of activated copper and column chromatography on neutral alumina in benzene, 2.0 g (15%)of authentic 12 was isolated. 2’-Chloro-4-Ylydroxy-4’-nitro-2,3,6-trimethylbiphenyl ( 1, R = H). A solution of 4.04 g (0.0103 mol) of benzyl ether 1, R = CH2CeH5, and 40 ml of 30-32% HBr in acetic acid was refluxed for 2.5 hr and then poured over 125 g of ice. The methylene chloride extract yielded 4.37 g of brown oil. Chromatography of a benzene solution by elution with hexane yielded 0.22 g (100%) of benzyl bromide; elution with 1:l benzene-methanol gave 2.65 g (88%) of crude product. Recrystallization from ethyl acetate-hexane three times gave tan crystals, mp 140.5-142.0’. Anal. Calcd for C15H14CIN03: C, 61.75; H, 4.84; N, 4.80. Found: C, 61.62; H, 4.91; N, 4.56. NMR (CDC13) 8 , 6 1.90 (6 H, 2,3-C&); 9, 2.22 (3 H, 6CH3); 9, 5.15 (1 H, OH); ~ , 6 . 6 3(1H, 5-H); d (J= 10 Hz), 7.29 (1 H, 6’-H); q (J = 2, 10 Hz), 8.17 (1H, 5’-H); d (J 2 Hz), 8.37 (1 H, 3’-H); ir (KBr) 3550 (OH); 1350, 1520 cm-* (NOz). 2‘-Chloro-4’-nitro-4-octyloxy-2,3,6-trimethylbiphenyl (1, R = n-CsH17). Method A. Iodo compound 9 (4.47 g, 0.0119 mol), 3.38 g (0.0119 mol) of 2-chloro-4-nitroiodobenzene(11): and 11.9 g of “activated” copper were allowed to react by the procedure given
J. Org. Chem., Vol. 41, No. 3, 1976
559
above for the preparation of 1, R = CHzCsH5. Elution of the crude product from neutral alumina with 1:l benzene-hexane gave 2.22 g of recovered 9 and 0.83 g of 1, R = n-CsH17. Elution with 2% methanol-benzene gave 0.96 g of symmetrical biphenyl 12. Based on unrecovered starting material the yield of 1, R = n-CsH17, was 53% and that of 12 67%. The conversion of 9 and 11 to 1, R = n-CsH17, was 17%. Upon standing 1, R = n-CsH17, crystallized. Several recrystallizations from chloroform-methanol by use of seed crystals led to off-white crystals, mp 59.5-60.5’. Anal. Calcd for C23H30ClN03: C, 68.39; H, 7.49; C1, 8.78; N, 3.47. Found C, 67.93; H, 7.26; C1, 9.07; N, 3.57. NMR (CDCl3) m, 6 0.6-2.24 [24 H, OC(CH2)s CH3 and CH3’s including 8, 1.77 (3-CH3), s, 1.83 (2CH3), S, 2.05 (6-CH3)I;t (J= 6 Hz), 3.73 (2 H, OCH2); 9, 6.24 (1 H, 5-H); d (J = 8 Hz), 6.84 (1 H, 6’-H); q ( J = 2, 8 Hz), 7.67 (1H, 5’H); d (J = 2 Hz), 7.85 (1 H, 3’-H); ir (neat) 1070, 1240, 1260, 1290 (ArOCH,); 1360,1560 cm-I (NOz). Method B.A solution of 5.00 g (0.0171 mol) of phenol I, R = H, 4.12 g (0.0171 mol) of n-octyl iodide, 1.26 g (0.0193 mol) of 87% potassium hydroxide, 30 ml of ethanol, and 20 ml of dioxane was refluxed for 22 hr, concentrated to -one-half volume, and diluted with water. The ether extract was washed with 10% sodium hydroxide, water, saturated salt solution, and water and dried over sodium sulfate. Evaporation gave 6.18 g (89%) of crude 1, R = nC8H17. 4~-Amino-2’-chloro-4-octyloxy-2,3,6-trimethylbiphenyl (2, R = o-CgH17). A mixture of 4.80 g (0.0119 mol) of nitro compound 1, R = n-CsH17, 10.7 g (0.0475 mol) of stannous chloride dihydrate, 30 ml of concentrated hydrochloric acid, and 90 ml of ethanol was refluxed for 18.5 hr, cooled, and poured into excess cold sodium hydroxide solution. The methylene chloride extract was washed with water, dried (sodium sulfate), and evaporated to afford 4.33 g (97%)of yellow oil. The oil was chromatographed on neutral alumina by elution with hexane and benzene-hexane mixtures. None of the fractions CMbe induced to crystallize. Anal. Calcd for C23H32ClNO: C, 73.8Rx, 8.63; C1, 9.48; N, 3.74. Found: C, 72.82; H, 8.30; C1, 8.72; N, 3.57. NMR (CDC13) m, 6 0.4-2.14 [(24 H, OC(CH2)&H3 and s, 1.80 (3-CH3), s, 1.84 (2-CH3), s, 2.02 (6CH3)]; broad s, 3.37 (2 H, ”2); t ( J = 6 Hz), 3.70 (2 H, OCHzCHz-); m, 6.05-6.87 (4 H, aromatic H); ir (neat) 1070, 1240, 1270, 1290 (ArOCHz); 3390,3480 cm-I (“2). 4’-(p-Cyanobenzylideneamino)-2’-chloro-4-octyloxy-2,3,6trimethylbiphenyl (13). The amine 2, R = n-CsH17, and an equimolar amount of p-cyanobenzaldehyde were heated in absolute ethanol in the presence of a drop of acetic acid for 4 hr while the ethanol was allowed to boil off. The residue was purified by recrystallization from hexane as pale yellow crystals, mp 96.5-98.5”. Anal. Calcd for C31H36CIN20: C, 76.44; H, 7.24; N, 5.75. Found: C, 76.13; H, 7.18; N, 5.63. Ir (KBr) 2250 cm-’ (CN); 1010, 1030, 1040, 1060,1240,1270 cm-I (ArOCH2);no C=O or NH. Ethyl 4-Iodo-3-nitrobenzoate (14, R = C2H5). Nitration of ethyl p-iodobenzoate1° with nitric acid-sulfuric acidlo afforded the desired compound, mp 78-81O (methanol), as orange crystals in 29% yield (reportedlo mp 88-89.5’, yield 85%): NMR (CDC13) t ( J . = 7 Hz), 6 1.44 (3 H, CH3); q, 4.45 (2 H, CHz); q ( J = 2,’s Hz), 7.86 (1 H, 6-H); d (J = 8 Hz), 8.17 (1 H, 5-H); d (J = 2 Hz), 8.41 (1H, 2-H). 4’-Carbethoxy-2‘-nitro-4-octyloxy-2,3,6-trimethylbiphenyl (3, R = CzHfi; R‘ = n-CsH1.r). An Ullmann reaction of 10.9 g (0.0291 mol) of 9,3.2 g (0.0100 mol) of 14, R = C2H5, and 5.52 g of activated copper by the method described above for the preparation of 1, R = CHzCsH5, and column chromatography on acidic alumina gave 0.33 g (7.5%)of crude 3, R = C2H5; R’ = n-CsH17 (1:l benzene-hexane elution). This oil could not be crystallized: NMR (CDC13) m, 6 0.5-2.14 [24 H, OC(CH&CH3 and s, 1.75 (3-CH3), s, 1.80 (2-CH3), 8, 2.00 (6-CH3)I; t ( J = 5 Hz), 3.67 (2 H, OCHzCH2); S, 6.19 (1 H, 5-H); d ( J 8 Hz), 6.91 (1 H, 5’-H); q (J = 2, 8 Hz), 7.87 (1 H, 6’-H); d ( J = 2 Hz), 8.14 (1 H, 3’-H); broad s, 9.59 (-COOH); ir (neat) 1370, 1530 (NOz); 1030, 1250, 1290, 1310 (ArOCH2); 1750 cm-1 (C=O). 4’-Carboxy-2’-nitro-4-octyloxy-2,3,6-trimethylbiphenyl (3, R = H R‘ = a-CsH17). A solution of 0.33 g (0.749 mmol) of the ester 3, R = R’ = n-CsH17,0.15 g of sodium hydroxide, 5 ml of dioxane, and 2 ml of water was refluxed for 2 hr and poured into cold dilute hydrochloric acid. The ether extract was washed with water, dried (sodium sulfate), and evaporated, leaving 0.28 g (90%) of crude product. Three recrystallizations from ethyl acetate-hexane gave yellow crystals, mp 175.6-176.1’. Anal. Calcd for Cz4H3iN05: C, 69.71; H, 7.56; N, 3.39. Found: C, 69.95; H, 7.48; N, 3.38. Ir (KBr) 1030,1090,1260,1295 (ArOCH,); 1360,1520 (NOz); 1700 (C=O); 3000-2500 cm-’ (OH).
560
J. Org. Chem., Vol. 41, No. 3, 1976
Notes
2,2’-Dibromo-6,6’-dichloro-4,4’-dinitrobiphenyl (4). Bromibased on ring opening of cyclopropyl ketones followed by nation of 12 by the method used by Harris and Mitchell” led to a oxidation, were reported.1° Most of these routes are either quantitative yield of 4 as light orange plates, mp 157.8-160.0° lengthy, require multistep preparation of a special reagent, (ethanol). Anal. Calcd for C1ZHldBr2C12N204: C, 30.61; H, 0.86; Br, or give low overall yields. 33.94. Found: C, 30.30; H, 0.91; Br, 34.70. Ir (KBr) 1360, 1530 cm-l’ We have found a short route which begins with readily (Nod. Attempted Self-coupling of 4-Benzyloxy-2,3,6-trimethylio- available materials, and gives high yields of y-keto aldedobenzene (8) via the Grignard Reagent. The reported procehyde, as outlined in Scheme I. Grignard reagents have dure for preparation of bimesityl13was applied to 8; the only product, formed in about 80% yield, was 3-benzyloxy-1,2,5-trimethylScheme I benzene (see below). 3-Benzyloxy-1,2,5-trimethylbenzene.Benzylation of the phe0 nol 6 by the procedure described above for the preparation of 8 led < ~ ~ C H 2 C H , M g B r CIC(CH2),CH, to a 75% yield of the ether as an oil. Elution through neutral alumina and distillation gave a colorless oil, bp 165O (2.7 mm), which crystallized (mp 34.5-37.5’). Anal. Calcd for C16H180: C, 84.91; H, 8.02. Found: C, 85.44; H, 7.97. NMR (CDC13) s, 6 2.13 (6 H, 1,2CH,); s, 2.20 (3 H, 5-CH3); s, 5.14 (2 H, OCH2); broad 8, 6.50 (2 H, 4,6-H); m, 7.2 (5 H, CsH5); ir (neat) 1120, 1225, 1250, 1290 cm-l 0 (ArOCH2-); no OH.
+ II
Registry No.-1 (R = C H ~ C E H ~57362-66-0; ), 1 (R = H), 57362-67-1; 1 (R = n-C8H17), 57362-68-2;2 (R = n-CsHlT), 5736269-3; 3 (R = CzH5; R’ = n-CgH17), 57362-70-6; 3 (R = H; R’ = nC8H17), 57362-71-7; 4, 57362-72-8; 6, 697-82-5; 7, 7282-02-2; 8, 57362-73-9; 9, 57362-74-0; 11, 41252-96-4; 12, 57362-75-1; 13, 57362-76-2; 14 (R = CzHj), 57362-77-3; benzyl chloride, 100-44-7; n-octyl chloride, 629-27-6; p-cyanobenzaldehyde, 105-07-7; ethyl p-iodobenzoate, 51934-41-9; 3-benzyloxy-1,2,5-trimethylbenzene, 57362-78-4. R e f e r e n c e s a n d Notes (1) H. W. Gibson, Can. J. Chem., 51,3065 (1973). (2) J. E. Kuder, H. W. Gibson, and D. Wychick, J. Ofg. Chem.. 40, 875 (1975). (3) H. W. Gibson and F. C. Bailey, Tetrahedron, 30, 2043 (1974). (4) H. W. Gibson and F. C. Bailey, Can.J. Chem., 53, 2162 (1975). (5)H. W. Gibson and F. C. Bailey, J. Chem. Soc.. Perkin Trans. 2, in press. (6) H. W. Gibson, J. Am. Chem. SOC.,97,3832 (1975). (7) E. L. Eiiel, “Stereochemistryof Carbon Compounds”,McGraw-Hill. New York, N.Y., 1962, p 156 ff: “Steric Effects in Conjugated Systems”, G, W. Gray, Ed., Academic Press, New York, N.Y., 1958. (8) H. W. J. Cressman and J. R. Thirtle, J. Org. Chem., 31, 1279 (1966). (9) L. M. Litvinenko, A. P. Grekov, N. N. Verkhovod, and V. P. Dzyuba, Zh. Obshch. Khim., 26, 2524 (1956): Chem. Abstr., 51, 5008e (1957). (10) N. E. Searie and R. Adams, J. Am. Chem. Soc., 55, 1649 (1933). (1 1) M. M. Harris and R. K. Mitchell, J. Chem. Soc., 1905 (1960). (12) R. B. Carlin, J. Am. Chem. Soc., 67,928 (1945). (13) W. W Mayer and R. Adams, J. Am. Chem. Soc., 51,630 (1929). (14) “Organic Syntheses”, Collect. Vol. II, Wiley, New York, N.Y., 1943, p 446.
A N e w y-Keto Aldehyde Synthesis John C. Stowell Department of Chemistry, University of New Orleans, Neui Orleans, Louisiana 70122 Received September 8,1975 y-Keto aldehydes are a n important class of compounds especially as intermediates for the preparation of cyclopenten0nes.l 1,4-Diketones have found wide application in the Paal-Knorr synthesis of pyrroles, furans, and thiophenes,2 as well as in pyridazine synthesis: which suggests that yketo aldehydes may be useful in the synthesis of these heterocycles as well. T h e available routes to y-keto aldehydes include ring opening of substituted furans: radical addition of an aldehyde to acrolein diethyl acetal,5 oxidative cleavage of olefins,6 alkylation of 2,4,4,6-tetramethyldihydrooxazine with 2-iodomethyl-l,3-dioxolane,7and alkylation of 2-ethoxyallyl vinyl sulfide followed by thio-Claisen rearrangement.8 Recently a route using condensation of the dianion of y-oxosulfone acetals with esters followed by removal of activating and protecting g r o u p ~ a, ~n d another
10
H,O o d i c acid
0
II
II
HCCH&H2C(CHJbCH3 found occasional use for preparation of ketones from acid chlorides b u t the yields are usually low owing t o reaction with the ketone, leading to tertiary alcohol^.^^-^^ Even a t dry ice temperature and using inverse addition, the yields are not improved.14 We have found that the Grignard reagent derived from 2-(2-bromoethyl)-1,3-dioxane is exceptional; it affords ketone in high yield with no more than a trace of tertiary alcohol at dry ice temperature. Grignard reagents derived from @-haloketals and acetals are known to be unstable15 owing to their tendency to undergo intramolecular attack leading initially to cyclopropyl ethers. Biichil6 and others17 have used the reagent derived from 2-(2-bromoethyl)-1,3-dioxolaneby preparing i t a t 30-35’. This reagent is destroyed if the solvent tetrahydrofuran ( T H F ) is allowed to reflux. We chose to use the sixmembered ring acetal and found that its greater stability allowed quick, high-yield preparation of the Grignard reagent a t reflux in THF. This then gave 2-(3-oxononyl)-1,3dioxane in 92% yield based on the bromo compound or the acid chloride. The six-membered-ring acetal apparently has greater equilibrium stability than the five-membered ring, since boiling the intermediate in aqueous acid gives only partial hydrolysis to the keto aldehyde. However, the equilibrium is easily displaced toward hydrolysis with removal of the product by continuous steam distillation as it is formed. This gave y-ketodecanal in 89% yield. This particular product is an intermediate in a synthesis of dihydrojasrnone.* T h e Grignard reagent derived from 2-(2-bromoethyl)4,4,6-trimethyl-1,3-dioxanealso gives ketones in better than 90% yields but the acetal is more stable and therefore less readily removed with aqueous acid. The usual technique for preventing overreaction is to first convert the Grignard to the zinc or cadmium reagent, which then affords ketones in better y i e l d ~ . ~In~ our J ~ earlier efforts, we prepared the zinc reagent directly from 2(2-iodoethy1)dioxolane at 51” in dimethylformamide (DMF);19 however, we were unable t o isolate any ketone products from reaction with acid chlorides. T h e formation of the organozinc iodide compounds could be followed in the NMR,20 e.g., that from the iodoacetal gave a high-field triplet a t 6 0.13 ppm, comparable with a quartet a t 6 0.01 obtained from ethyl iodide and Zn in DMF. T h e Grignard
